---
title: "Data Science Salaries Dashboard"
output: 
  flexdashboard::flex_dashboard:
    orientation: columns
    source_code: embed
    vertical_layout: scroll
    theme: flatly
    logo: "C:/Users/nishn/Downloads/icons8-money-with-wings-48.png"
---

<style>                     
.navbar {
  background-color:purple;
  border-color:white;
}
.navbar-brand {
color:black!important;
}

</style>  


```{r setup, include=FALSE}
library(flexdashboard)
library(dplyr)
library(ggplot2)
library(DT)
library(plotly)
library(wordcloud2)
library(sf)
library(spData)
```
  
Data Table
===================
```{r}
data=read.csv("C:/Users/nishn/Downloads/ds_salaries.csv")
data<- data %>%
  mutate(employment_type = case_when(
    employment_type == "FT" ~ "Full Time",
    employment_type == "PT" ~ "Part Time",
    employment_type == "CT" ~ "Contract",
    employment_type == "FL" ~ "Freelance",
    TRUE ~ employment_type
  ))
data<- data %>%
  mutate(experience_level = case_when(
    experience_level == "EN" ~ "Entry Level/Junior",
    experience_level == "MI" ~ "Mid-level/Intermediate",
    experience_level == "SE" ~ "Senior-level/Expert",
    experience_level == "EX" ~ "Executive-level/Director",
    TRUE ~ experience_level
  ))
data<- data %>%
  mutate(company_size = case_when(
    company_size == "L" ~ "Large",
    company_size == "M" ~ "Medium",
    company_size == "S" ~ "Small",
    TRUE ~ company_size
  ))
data$remote_ratio=as.character(data$remote_ratio)
data<- data %>%
  mutate(remote_ratio = case_when(
    remote_ratio == "0" ~ "No Remote Work",
    remote_ratio == "50" ~ "Partially Remote",
    remote_ratio == "100" ~ "Fully Remote",
    TRUE ~ remote_ratio
  ))



datatable(data)

```

Designation Analysis
======================

Column {data-width=500}
-----------------------------------------------------------------------

### Job Designation and Experience Level

```{r}
top_job_titles <- data %>%
  group_by(job_title) %>%
  summarise(count = n()) %>%
  arrange(desc(count)) %>%
  slice_head(n = 15) %>%
  pull(job_title)

# Filter the original data for the top 10 job titles
filtered_data <-data %>%
  filter(job_title %in% top_job_titles)

# Create a clustered bar chart with ggplot for the top 15 job titles and experience levels
clustered_bar_chart_top15 <- ggplot(filtered_data, aes(x = job_title, fill = experience_level)) +
  geom_bar(position = "dodge", stat = "count") +
  labs(title = " Experience Level vs. Top 15 Job Designations",
       x = "Job Title",
       y = "Count",
       fill="Experience Level") +
  theme_minimal()+
  theme(axis.text.x = element_text(angle = 45, hjust = 1 ),legend.position = "bottom",
        plot.title = element_text(hjust = 0.5))

# Convert ggplot to plotly for the top 15 job titles
interactive_clustered_bar_top15 <- ggplotly(clustered_bar_chart_top15, tooltip = "all")
interactive_clustered_bar_top15
```


### Job Designation - Word Cloud

```{r,fig.width=15}
text=data$job_title
my_data <- data.frame(text = text, freq = 1, stringsAsFactors = FALSE)
my_agr <- aggregate(freq ~ ., data = my_data, sum)
wordcloud2(data = my_agr,size=5)
```





### Job Designation and Salary

```{r}
violin_plot <- ggplot(filtered_data, aes(x = job_title, y = salary, fill = job_title)) +
  geom_violin() +
  labs(title = "Salary vs. Top 15 Job Titles",
       x = "Job Title",
       y = "Salary",
       fill="Job Title") +
  scale_y_continuous(limits=c(0,1000000))+
  theme_minimal()+
  theme(axis.text.x = element_text(angle = 45, hjust = 1 ),legend.position = "right",
        plot.title = element_text(hjust = 0.5))

ggplotly(violin_plot)

```
Salary Analysis
==================

Column {data-width=500}
-----------------------------------------------------------------------
### Salary Distribution
```{r}
gg=ggplot(data, aes(x = salary_in_usd)) +
  geom_density(alpha = 0.7,color="blue") +
  geom_rug(color="red") +  # Add a rug plot below
  labs(title = "Salary Distribution", x = "Salary") +
  theme_minimal()+
  scale_x_continuous(labels = scales::number_format(scale = 1))+
  theme(plot.title = element_text(hjust = 0.5))
  


ggplotly(gg)
```



### Salary Distribution based on Company Size
```{r}

gg <- ggplot(data, aes(x = salary_in_usd, color = company_size)) +
  geom_density(alpha = 0.7) +
  labs(title = "Salary Distribution based on Company Size", x = "Salary", color = "Company Size") +
  theme_minimal()+
  scale_x_continuous(labels = scales::number_format(scale = 1))

# Convert ggplot object to a plotly object
plotly_chart <- ggplotly(gg)

# Customize the appearance of the plotly chart
plotly_chart <- plotly_chart %>%
  layout(title = "Salary Distribution based on Company Size", xaxis = list(title = "Salary"), yaxis = list(title = "Density"))

# Display the interactive plot
plotly_chart
```

Column {data-width=500}
-----------------------------------------------------------------------

### Median Salary by Experience Level and Remote Ratio
```{r}
grouped_data1 <- data %>%
  group_by(experience_level, remote_ratio) %>%
  summarise(median_salary = median(salary))

plot_ly(data = grouped_data1, x = ~experience_level, y = ~remote_ratio, z = ~median_salary, type = 'heatmap') %>%
  layout(title = "Heatmap of Median Salary by Experience Level and Remote Ratio",
         xaxis = list(title = "Experience Level", categoryorder = "array", categoryarray = c("Entry Level/Junior", "Mid-level/Intermediate", "Senior-level/Expert", "Executive-level/Director")),
         yaxis = list(title = "Remote Ratio",categoryorder="array",categoryarray=c("Fully Remote","Partially Remote","No Remote Work")),
        legend=list(title="Median Salary"))

```

### Salary Distribution based on Experience
```{r}
gg <- ggplot(data, aes(x = salary_in_usd, color = factor(experience_level))) +
  geom_density(alpha = 0.8) +
  labs(title = "Salary Distribution based on Experience", x = "Salary", color = "Experience") +
  theme_minimal()+
  scale_x_continuous(labels = scales::number_format(scale = 1))

# Convert ggplot object to a plotly object
plotly_chart <- ggplotly(gg)

# Customize the appearance of the plotly chart
plotly_chart <- plotly_chart %>%
  layout(title = "Salary Distribution based on Experience", xaxis = list(title = "Salary"), yaxis = list(title = "Density"))

# Display the interactive plot
plotly_chart
```




Max & Min Salary Analysis
====================
Column {data-width=500}
-----------------------------------------------------------------------
### Median Salary for Top 10 Highest Paid Job Titles

```{r}

top_salary <- data %>%
  group_by(job_title) %>%
  summarise(med_salary = median(salary_in_usd)) %>%
  arrange(desc(med_salary)) %>%
  head(10)

top_10=top_salary%>% 
    ggplot(aes(x = reorder(job_title, -med_salary), y = med_salary, fill = job_title,text = paste("Job Title: ", job_title, "<br>Median Salary: $", round(med_salary, 2)))) +
  geom_bar(stat = "identity") +
  labs(title = "Median Salary for Top 10 Highest Paid Job Titles",
       x = "Job Title",
       y = "Salary") +
  theme_bw()+
  theme(axis.text.x = element_text(angle = 45, hjust = 1),legend.position = "none",
        plot.title = element_text(hjust = 0.5))+  scale_y_continuous(labels = scales::number_format(scale = 1))
  

ggplotly(top_10,tooltip = "text")
```


### Bubble Chart of Job Title v/s Experience Level by Salary (Highest Paid)
```{r}
med_salaries_high <- 
  inner_join(data,top_salary,by="job_title") %>%
  group_by(job_title, experience_level) %>%
  summarise(med_salary = median(salary_in_usd))

theme_set(theme_bw())  # pre-set the bw theme.
g <- ggplot(med_salaries_high, aes(job_title, experience_level,
    text=paste("Job Title:",job_title, "<br>Experience Level:",experience_level, "<br>Salary:",round(med_salary,2),"$"))) + 
  labs(title="Job Title v/s Experience Level by Salary (Highest Paid)",
       x="Job Title",
       y="Experience Level")+
  geom_jitter(aes(col=experience_level,size=med_salary))+
  theme(axis.text.x = element_text(angle = 45, hjust = 1),
        plot.title = element_text(hjust = 0.5),legend.position = "none")
ggplotly(g,tooltip = "text")

```



Column {data-width=500}
-----------------------------------------------------------------------
### Median Salary for Top 10 Lowest Paid Job Titles

```{r}
bottom_salary <- data %>%
  group_by(job_title) %>%
  summarise(med_salary = median(salary_in_usd)) %>%
  arrange(med_salary) %>%
  head(10)
  
bottom_10=bottom_salary%>% 
  ggplot(aes(x = reorder(job_title, med_salary), y = med_salary, fill = job_title,text = paste("Job Title: ", job_title, "<br>Median Salary: $", round(med_salary, 2)))) +
  geom_bar(stat = "identity") +
  labs(title = "Median Salary for Top 10 Lowest Paid Job Titles",
       x = "Job Title",
       y = "Salary") +
  theme_bw()+
  theme(axis.text.x = element_text(angle = 45, hjust = 1),legend.position = "none",
        plot.title = element_text(hjust = 0.5))+  scale_y_continuous(labels = scales::number_format(scale = 1))

ggplotly(bottom_10,tooltip = "text")

```


### Bubble Chart of Job Title v/s Experience Level by Salary (Lowest Paid)
```{r}
med_salaries_low <- 
  inner_join(data,bottom_salary,by="job_title") %>%
  group_by(job_title, experience_level) %>%
  summarise(med_salary = median(salary_in_usd))

theme_set(theme_bw())  # pre-set the bw theme.
g <- ggplot(med_salaries_low, aes(job_title, experience_level,
    text=paste("Job Title:",job_title, "<br>Experience Level:",experience_level, "<br>Salary:",round(med_salary,2),"$"))) + 
  labs(title="Job Title v/s Experience Level by Salary (Lowest Paid)",
       x="Job Title",
       y="Experience Level")+
  geom_jitter(aes(col=experience_level,size=med_salary))+
  theme(axis.text.x = element_text(angle = 45, hjust = 1),
        plot.title = element_text(hjust = 0.5),legend.position = "none")
ggplotly(g,tooltip = "text")
```


Work Year Analysis
==========================
Column {data-width=500}
-----------------------------------------------------------------------
### Scatter Plot of Median Salary by Year
```{r}
med_salaries<- 
  data %>% 
  group_by(work_year,experience_level) %>%
  summarise(med_salary = median(salary_in_usd))

scatter_plot<- ggplot(med_salaries,aes(x =work_year, y = med_salary, 
  color = experience_level,group=experience_level,
  text=paste("Year:",work_year,"<br>Salary:",round(med_salary,2),"$",
             "<br>Experience Level:",experience_level))) +
  geom_point(size = 3) +
  geom_line(size = 1, aes(color = experience_level))+
  labs(title = "Scatter Plot of Median Salary by Year",
       x = "Year",
       y = "Salary")+
  theme(plot.title = element_text(hjust = 0.5))

ggplotly(scatter_plot,tooltip="text")
```

### Remote Ratio by Work Year
```{r}
percentage_data <- data %>%
  group_by(work_year, remote_ratio) %>%
  summarise(count = n()) %>%
  group_by(work_year) %>%
  mutate(percentage = count / sum(count) * 100)

# Create a stacked bar chart for remote ratio by work year (interactive with percentage on hover)
gg <- ggplot(percentage_data, aes(x = work_year, y = percentage, fill = remote_ratio)) +
  geom_bar(stat = "identity") +
  labs(title = "Remote Ratio by Work Year", x = "Work Year", y = "Percentage",
       fill="Remote Ratio") +
  scale_y_continuous(labels = scales::percent_format(scale = 1)) +  # Format y-axis as percentage
  theme_minimal()+
  theme(plot.title = element_text(hjust = 0.5))


# Convert ggplot object to a plotly object
plotly_chart <- ggplotly(gg, tooltip = c("work_year", "remote_ratio", "percentage"))

# Display the interactive plot
plotly_chart

```


Column {data-width=500}
-----------------------------------------------------------------------
### Percentage Distribution of Work Years
```{r}
average_salaries<- 
  data %>% 
  group_by(work_year,company_size) 

percentage_data <- data %>% 
  group_by(work_year) %>% 
  summarize(count = n()) %>% 
  mutate(percentage = count / sum(count) * 100)

plot_ly(percentage_data, labels = ~work_year, values = ~percentage, type = "pie",
        hoverinfo = "label+percent") %>%
  layout(title = "Percentage Distribution of Work Years")

```


### Company Location by Work Year
```{r}
top_locations <- data %>%
  group_by(company_location) %>%
  summarise(location_count = n()) %>%
  arrange(desc(location_count)) %>% 
  head(15)

ij=inner_join(top_locations,data,by="company_location")

ij <- left_join(ij, world %>% select(iso_a2, name_long), by = c("company_location" = "iso_a2")) %>%
  mutate(company_location = coalesce(name_long, company_location)) %>%
  select(-name_long) %>% 
  st_drop_geometry(geom)



percentage_data1<- ij %>%
  group_by(work_year, company_location) %>%
  summarise(count = n()) %>%
  group_by(work_year) %>%
  mutate(percentage = count / sum(count) * 100)

# Create a stacked bar chart for remote ratio by work year (interactive with percentage on hover)
gg <- ggplot(percentage_data1, aes(x = work_year, y = percentage, fill = company_location)) +
  geom_bar(stat = "identity") +
  coord_flip()+
  labs(title = "Company Location by Work Year", x = "Work Year", y = "Percentage",
       fill="Company Location") +
  scale_y_continuous(labels = scales::percent_format(scale = 1)) +  # Format y-axis as percentage
  theme_minimal()+
  theme(plot.title = element_text(hjust = 0.5))

# Convert ggplot object to a plotly object
plotly_chart <- ggplotly(gg, tooltip = c("work_year", "company_location", "percentage"))

# Display the interactive plot
plotly_chart

```


Location Analysis
=====================

Column {data-width=500}
----------------------------
### Median Salary by Company Location

```{r map, echo=FALSE, results='asis'}

world <- spData::world


median_salary_by_country <- data %>%
  group_by(company_location) %>%
  summarise(med_salary = median(salary_in_usd)) %>% 
  arrange(desc(med_salary))

merged_data <- merge(world, median_salary_by_country, by.x = "iso_a2",
                     by.y = "company_location",all=TRUE) 

jmd=inner_join(merged_data,median_salary_by_country,by=c("iso_a2"="company_location")) %>% 
  select(iso_a2,name_long,med_salary.x) %>% 
  st_drop_geometry()


plot_ly(data = jmd, x = ~reorder(name_long, -med_salary.x), y = ~med_salary.x,
        type = 'bar', marker = list(color = rainbow(nrow(jmd)))) %>%
  layout(title = "Median Salary by Country", xaxis = list(title = "Country"), yaxis = list(title = "Median Salary"))

```



### Top Employee Residence countries and their company location

```{r}
emp_data=data%>% 
  group_by(employee_residence) %>% 
  summarise(count=n()) %>% 
  arrange(desc(count)) %>% 
  head(10)

data_top_emp=merge(data,emp_data,by="employee_residence",all=TRUE)
data_top_emp <- data_top_emp %>%
  filter(!is.na(count)) %>% 
  group_by(employee_residence)

dataind=data_top_emp %>% 
  filter(company_location!=employee_residence) %>% 
  select(company_location,employee_residence) %>% 
  group_by(company_location,employee_residence) %>% 
  summarise(count=n()) 

world=world %>% st_drop_geometry()

dataind <- left_join(dataind, world %>% select(iso_a2, name_long), by = c("company_location" = "iso_a2")) %>%
  mutate(company_location = coalesce(name_long, company_location)) %>%
  select(-name_long) %>% 
  st_drop_geometry(geom)

dataind <- left_join(dataind, world %>% select(iso_a2, name_long), by = c("employee_residence" = "iso_a2")) %>%
  mutate(employee_residence = coalesce(name_long, employee_residence)) %>%
  select(-name_long)%>% 
  st_drop_geometry()

# Create a ggplot object
p <- ggplot(dataind, aes(x = reorder(employee_residence, -count), text=paste("Company Location:",company_location,"<br>Count:",count), y = count, fill = dataind$company_location)) +
  geom_bar(stat = "identity") +
  labs(title = "Stacked Bar Chart of Counts by Employee Residence and Company Location",
       x = "Employee Residence",
       y = "Count",
       fill="Company Location") +
  theme_minimal()+
  theme(plot.title = element_text(hjust = 0.5))
  

# Convert ggplot object to plotly
p <- ggplotly(p,tooltip ="text")

# Show the interactive plot
p
```